© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 72789
    Prev
    Next
    INDEX
    Explanations

    Spinner, assembly, Hutton"}, k

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron fires on unusually long, multi-character tokens—i.e. very long words or identifiers without spaces.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    assume
    -0.84
    Eid
    -0.81
     Wählen
    -0.81
    IAm
    -0.80
    Vikings
    -0.79
     sacrificed
    -0.78
    SNOW
    -0.76
     noël
    -0.76
     eigenlijk
    -0.75
     robusto
    -0.73
    POSITIVE LOGITS
     with
    0.89
     when
    0.83
     until
    0.79
    XmlAccessType
    0.78
    tree
    0.77
    ンダル
    0.75
    apkan
    0.75
    When
    0.74
    Most
    0.73
    ਦੀ
    0.73
    Activations Density 0.025%

    No Known Activations